Lustra

What it does

Lustra ingests raw legislative bills from US and Polish government APIs, processes them through language models to remove political framing, and summarizes the content in plain language. Users then vote on which bills matter most. The platform displays legislation in order of community interest rather than editorial selection.

Who it is for

The platform targets people who want to understand legislative changes without media interpretation. This includes civic-minded citizens, policy researchers, and anyone tracking how laws might affect their work or interests. The stated problem is that 95% of legislation goes unnoticed because legal texts are difficult to parse and media coverage prioritizes outrage over substance.

Pricing

The site does not show prices.

How it stands out

Lustra uses a community voting mechanism called "Shadow Parliament" to rank bills by genuine user interest rather than editor judgment. The technical approach strips political language from source documents using language models set to deterministic outputs (temperature=0, strict JSON formatting) to minimize subjective interpretation. The platform accepts both PDF and XML formats from government sources, suggesting it handles messy real-world data. It operates across multiple jurisdictions, beginning with US and Polish legislation, which could differentiate it from US-only competitors.

What a founder should check

First, examine whether existing news aggregators, legislative tracking services like Congress.gov, or specialized lobbying platforms already serve this need cheaply or free. Many jurisdictions publish their own bill summaries; understand what problem Lustra solves beyond aggregation and formatting.

Second, validate the switching costs. Prospective users may already track bills through alerts from their representatives, legal databases their organizations subscribe to, or news services they trust. A founder should test whether a community voting system actually improves on algorithmic ranking or editorial curation, or whether users prefer predictability.

Third, test the moat. The language model approach to removing political bias is reproducible by well-funded incumbents (news organizations, government agencies, legal platforms). The community voting mechanism could be copied. The real defensibility lies in network effects—whether more users create more valuable rankings, and whether that advantage compounds. A solo founder should validate whether the platform achieves sufficient scale to make the voting signal meaningful before larger competitors enter.